Identifying direct protective factors for nonviolence

Background: The CDC recently organized a panel to examine whether a series of constructs consistently acted as risk and/or direct protective factors for youth violence across four longitudinal studies. Analyses first examined constructs commonly assessed across all four studies and then included constructs unique to each study. Purpose: This paper describes findings from the Pittsburgh Youth Study (PYS) as part of this supplement to the American Journal of Preventive Medicine documenting the findings from the project. Methods: Participants were boys in the youngest cohort of the PYS (N=503), which was initiated in 1987-1988. Constructs measured at age 12 years were trichotomized to test whether they acted as risk and/or direct protective factors in predicting violence (i.e., assault, rape, robbery) across ages 13-14 years and 15-18 years. Results: Multivariate logistic regressions with predictors present across studies indicated that depressed mood (OR=1.96) and low religious observance (OR=1.88) were risk factors for violence at ages 13-14 years, whereas peer delinquency acted as both a risk (OR=2.34) and direct protective factor (OR=0.44). Low peer delinquency was also a direct protective factor (OR=0.41) for violence at ages 15-18 years. Analyses including predictors specific to the PYS indicated that negative attitude toward delinquency (OR=0.50) was protective against violence at ages 13-14 years, whereas the risk factors of low perceived likelihood of being caught (OR=1.81) and high neighborhood disorder/crime (OR=1.77) predicted violence at ages 15-18 years. Conclusions: Some factors may be best conceptualized as direct protective factors for nonviolence, whereas other constructs act primarily as risk factors that increase the probability of adolescent violence.